The current project proposes to construct a corpus of discourse production of Chinese speakers with aphasia, one of the underrepresented minority groups as listed by the National Institutes of Health (NIH). This database will encompass information on distinctive linguistic properties in Chinese aphasia, prosodic features of the Chinese aphasic speech, and non- verbal behaviors of Chinese speakers with aphasia as a result of stroke. The establishment of the corpus will provide the necessary foundation for aphasiologists and clinicians for research investigations into theoretical and clinical issues related to acquired language disorders in Chinese. The overarching goal of this study is to improve the planning of assessment and remediation procedures for Chinese speakers with aphasia worldwide, including those living in North America. This can be achieved by actively sharing the major deliverables of this project with any clinicians and researchers who work with Chinese speakers with acquired language deficits, i.e., a standardized assessment procedure, a multimedia database, and a multi-level coding system. A four-year investigative project that will be carried out in Hong Kong, a city with a linguistically homogenous Chinese-speaking community in China, is proposed. In Stage 1 of this project (12 months), a protocol for language testing and speech sample elicitation that is linguistically and culturally appropriate for Chinese speakers will be developed. Normative data from 180 native normal speakers in Cantonese will be collected. In Stage 2 (15 months), 180 native aphasic speakers in Cantonese will be tested. The aphasic language samples, coded in internationally-accepted transcription and tagging systems, will serve as an empirical basis for linguistic investigations of Chinese aphasiology. Stage 3 of the project (21 months) will involve development of a multi-modal and multi-level assessment tool for guiding diagnosis and intervention planning in Cantonese aphasia. In addition, statistical analyses of language-impaired and normative data will be carried out to explore a new classification system that considers verbal and non-verbal behaviors of aphasic narrative in Cantonese. The database that is constructed in the present proposed study will be the first of its kind in Asian languages for studying aphasia. It will be significant in the terms of both theoretical and clinical aspects. In terms of theoretical aspects, the project will facilitate linguists, psychologists, neuro-linguists, and related professionals or scholars to get access to this database for research purposes. The Chinese data will allow for cross-linguistic comparisons and enrich theory generalizations of post-stroke aphasia. With reference to the multi-modal linguistic and non-linguistic empirical variables measured in this study, new approaches to classify aphasic speakers can also be developed. In terms of clinical aspects, this database can serve as a valuable resource of clinical education for speech and language pathologists and related medical health-care professionals. The observation and analysis of multi-modal linguistic and non-linguistic behaviors in aphasia can lead to the development of a new assessment tool for analyzing aphasic narrative in Cantonese and diagnosing aphasia types. Finally, the multi-level communication measures of verbal and non-verbal behaviors proposed in this project will be available for guiding treatment design, selecting and prioritizing treatment targets, and monitoring treatment efficacy and efficiency systematically.